What is Alpha?

Positive alpha is the manager adding something the index didn't hand them. It is a residual, so it is only as meaningful as the benchmark fit — check R² before believing an alpha figure.

The model behind it is a single regression: the fund's excess return (over the 6.5% risk-free rate) against the benchmark's excess return, over the trailing three years of daily data. The slope of that line is beta; alpha is the intercept, annualized. It answers one question — given how much market exposure this fund carried, how much did it return on top of what that exposure alone would have earned? A fund with beta 1.1 in a year the index made 15% was 'owed' roughly 6.5% + 1.1 × (15% − 6.5%) ≈ 15.9%; if it made 18%, its alpha is about +2%.

Why R² matters first. Alpha is whatever the model could not explain, so if the benchmark explains little of the fund's movement — R² below about 0.7 — the intercept is mostly noise and a large alpha in either direction is not evidence of skill. A thematic or sector fund measured against a broad index will often show a big alpha with a poor fit; the honest reading is 'wrong benchmark', not 'great manager'. Every benchmark-relative figure on this site is computed against the closest total-return index for the fund's SEBI category, over the last 36 complete months, and the fit is shown beside it so this check is one glance.

Alpha and cost. The figure is computed from NAV, which is already net of the expense ratio, so a Regular plan's alpha is lower than its Direct twin's by roughly the distributor commission — the same portfolio, a different intercept. Compare Direct against Direct. And remember it is a trailing average: a manager who added 3% a year for two years and lost 2% in the third shows a modest positive alpha that hides the swing. Read it with the rolling-returns chart, which shows the same edge year by year.

For the formula and the constants behind this figure, see Methodology.